Thomas Wolf, co-founder of Hugging Face, is charting new paths in robotics. Open source is becoming the vector of this ambition. Through accessible solutions, he aims to open the door to a collective mastery of robotic technologies.
The democratization of robotics is at the heart of his vision. Uniting enthusiasts and experts to create an ecosystem focused on innovation. The library LeRobot, resulting from a common aspiration, piques the interest of researchers and developers.
The projects of Hugging Face promise to transform this sector by making it more accessible.
Hugging Face’s ambitions in robotics
Hugging Face, co-founded by Thomas Wolf, has established itself as an indispensable player in the field of AI. Now, the startup is turning to robotics, striving to make it accessible to all through open source. Thomas Wolf, as Chief Scientist, presents a clear vision: to democratize robotics by promoting collective innovation.
A platform serving diverse communities
Originally designed for researchers, the Hugging Face platform has expanded its reach, welcoming a diversity of users, from independent developers to large corporations. Today, anyone interested in AI can find value on this platform, which offers models, datasets, and applications accessible to all. More than 100,000 applications, referred to as “spaces,” are available.
A solid business model
The majority of the platform’s features remain free. Thomas Wolf emphasizes that the startup does not plan to restrict access to these options. Generating revenue through consulting with major clients, such as Amazon and Nvidia, the company has established a stable business model. This model relies on a combination of consulting, the Enterprise Hub offer, and individual subscriptions.
Investment in open-source robotics
The recent acquisition of the Bordeaux-based startup Pollen Robotics demonstrates Hugging Face’s ambitions in robotics. This integration aims to develop software solutions and create a standard library focused on robot control policies. The open-source library LeRobot has already been launched, seeking to structure datasets and gather researchers and enthusiasts.
Creation of affordable robots
Two open-source robots, HopeJR and Richie Mini, were recently unveiled. HopeJR, a humanoid, is offered at a price of 3,000 euros, while Richie Mini is available for 300 euros. These initiatives aim to democratize robotics, particularly for developers. These robots allow for learning, creating, and experimenting without bearing prohibitive costs.
Collaboration with the community
The community plays a crucial role in the development of Hugging Face projects. The creation of the SO-100, a 3D-printable robotic arm, comes from a community member. By publishing the plans as open source, Hugging Face facilitates the fabrication of this arm at low cost, with the price estimated at under 100 euros. The advanced version, SO-101, is planned to cost less than 500 dollars.
Toward an open future in robotics
The founder envisions a future where the robotics ecosystem will grow by sharing resources and data. The goal is to enable model transfer between different robots, which remains a challenge today. The standardization of datasets, through LeRobot, promises to revolutionize the sector, paving the way for better collaboration between researchers and entrepreneurs.
Future development of Hugging Face
Hugging Face will continue to invest in small models, a segment overlooked by large tech groups. The commitment to share recipes, datasets, and tools will help enrich the ecosystem. Thomas Wolf reaffirms the belief that sharing promotes innovation and expands possibilities, both for the AI sector and for robotics.
Frequently Asked Questions
What are the reasons behind Hugging Face’s ambition to democratize robotics?
Hugging Face aims to make robotics more accessible to a wide audience, particularly to developers and enthusiasts, by integrating AI and using open-source solutions that foster innovation.
How does Hugging Face plan to achieve its goals in robotics?
The company focuses on developing standardized libraries, like LeRobot, to facilitate dataset sharing and create algorithms that can be used by robots.
What is the LeRobot project launched by Hugging Face?
LeRobot is an open-source library that standardizes datasets to make robotics accessible, thereby encouraging contributions from researchers and enthusiasts.
What types of robots does Hugging Face offer and who is their target audience?
Hugging Face has launched robots such as HopeJR and Richie Mini, available at competitive prices, primarily targeting software developers and hobbyists.
How can contributors participate in Hugging Face’s open-source robotics project?
Contributors can join the LeRobot community, engage in improving datasets and algorithms, and discuss practices via the Discord platform set up by Hugging Face.
Why does Hugging Face opt for lighter robotics models?
Thomas Wolf and his team believe that more compact models, which can operate locally, offer significant potential while being more practical for various applications.
Does Hugging Face plan to have commercial products in the robotics sector in the future?
While the company does not consider robotics as its core activity, it is contemplating the possibility of diversifying its revenues through the sale of affordable robots in the future.
What advantages does the use of open source offer for Hugging Face in developing robotics?
Open source allows Hugging Face to share resources and knowledge, facilitate collaboration, and innovate more rapidly by mobilizing the community.
Is Hugging Face working on robotics projects in collaboration with other companies?
Yes, Hugging Face collaborates notably with Pollen Robotics, leveraging their respective expertise in hardware and software to enrich the robotics ecosystem.
How does Hugging Face ensure the quality and safety of the robots developed?
Quality and safety are ensured through an engaged community that tests and improves the proposed designs, as well as through an open-source development framework that promotes transparency.